Before this was a "feature phone" the Nokia 9000, i loved it.
When this was out I was using the OG Motorola Droid. I wanted one of these but they were not available on my carrier so I stayed on the Droid line. The hot-dog slide keyboard was amazing. I had full terminal control from these.
Nowadays I actually still rock a 2017 iphone SE; dimensions wise I don't want anything bigger in width or height, but i would gladly add some thickness to get a full tactile keyboard again.
I used an N900 for a few years (got one with the early adopter discount when it was released). While the keyboard was miles ahead of all the on-screen stuff at the time I don't think I want to go back to that keyboard from a current decent touch-screen phone. The keys were a bit too small for fast typing...
I wonder how it compared to the keyboard on the N97. Though it was a symbian based phone, the putty client allowed you to connect to a server and attach to a screen or tmux session, which worked okay with that keyboard and a few on-screen buttons.
The key size is alright, I'm able to type on it much faster than on a touchscreen - provided that I don't need to use anything else than regular letters. The most painful thing on that keyboard is lack of 4th row of keys.
Still have this little linux phone. It missed compass hw, so map experience was not great and also resistive display was inferior to iPhone, but still loved it. It's bad Nokia didn't manage it create viable platform, but rather switched to Windows phones and also died there with broken teeth.
They had released the N9 and there was a developer only N950. I'm still using my N9 even today though the lock mechanism on the SIM holder no longer works.
Wow, I had to abandon my N9 after all of my fixes including buying a new SIM holder failed. I still have my N9 and my N900. I really wish Nokia had stuck it out with the N9 OS a bit longer. It easily beat all of the Android phones at that time. I would still be using it if the CAs hadn't expired and the SIM holder wasn't horked.
I'm still using my N9, and I'm holding my SIM card in place using a big rubber band wrapped twice along the longer side of the phone :)
Lately the "main" button has been failing, and since tapping the screen doesn't get it out of sleep anymore, I can't pick up calls anymore. So to access the phone, I get out the SIM and re-insert it.
It's time to abandon it here too, but it was an excellent phone in use for almost 10y.
I don't look forward to the day I have to completely abandon it. I have mine set to show the screen when double tapping it (which still works better than my Nokia 7.2 Android phone).
> since tapping the screen doesn't get it out of sleep anymore, I can't pick up calls anymore.
Mine will show the the screen and prompt me to swipe up to answer the call. Does yours work in a different way, or is the touch system failing?
I thought long about buying another N9 second-hand, but in the meanwhile MeeGo/Harmattan has been abandoned, and I never succeeded with installing the Mer [0] project (revived Maemo/MeeGo successor). I'm using my phone regularly to program and run simple applications for my own use. For this I've used Tclkit [1] a lot on the N9 (and the N770).
I'm now more thinking about continuing that using AndroWish [2] instead (Tcl/Tk on Android).
My N9 in better times did exactly that when someone called, but now the screen stays black and nothing reacts (it still rings though). I don't know what causes this, because out of the sleep mode then touchscreen works fine. I have to pull out the SIM everytime as such, put it back in and log back into the network (and then often call back myself ofcourse ;)). Not going to bother anymore, but I suspect a software problem. It also started to crash/hang it's applications from time to time.
> I really wish Nokia had stuck it out with the N9 OS a bit longer
I think we have Stephan Elop to thank for that. I never really understood why Nokia didn't just focus on the market outside the US. Had they continued producing phones like the N9, I'm sure they could have retained a lot of market share in Europe, Africa, Middle East and Asia.
> I would still be using it if the CAs hadn't expired
I'm not sure whether all of the CAs had expired, but I wonder if it would be possible to get an updated package with up to date CAs. The main issue I have is that it can't negotiate an agreed upon cipher when trying to establish a TLS connection.
I'm still able to use the Firefox browser to browse a number of websites like Facebook and Reddit. Hacker news still works with the built-in browser. Facebook sometimes works with the default browser and the old subdomain reddit website will lock up the default browser in my experience (the current reddit website layout is a no-go).
One nice thing about the resistive screen is that it's pressure-sensitive. This was particularly great for painting apps - I remember using one called MyPaint which mapped touch pressure to the intensity of your brush stroke, which let me draw things that just weren't possible on iPhones of the time.
A family friend worked as an outside something at Nokia HQ around 2007(?) when the iPhone was announced.
Apparently there was an ongoing joke inside of the company that every button on these phones was one of the executives. And so you had multiple buttons/ways of doing one thing because consensus. Sounds like they were just bogged down and crippled in a corporate way that so many European companies were between 1990-2020.
Apparently when the iPhone was announced the whole company was in deep shock and disarray. They were planning on the first touchscreen phones hitting the market around 2010-2015, and here was a company promising to deliver one by 2008.
Because the iPhone was an AT&T exclusive in the US, Verizon pressured Blackberry to rush the release of their first touchscreen device, the BB Storm. It was released as a Verizon exclusive in late 2008, and coincided with the release of HTC/Tmobile G1, the first Android phone.
BB tried to argue that BBOS6 was a click trackball based OS and would not transition to touch well. In the end they went for a bizarre hybrid design where the screen pushed down when you tapped something on the screen. Like a button would. I got used to it over time but the first time I used it, I thought I'd dislocated the screen.
Within a year, Verizon realized it wasn't going to rival AT&T w/ BB, and had Motorola create the Droid brand, which may have been one of the most popular Android phones in the US before Samsung arrived.
Webpage would render incredibly slowly because Blackberry weren't capable of putting a wifi chip in the handset (it was limited to edge if I recall). The whole UI was sluggish. Cherry on top? No apps. There was a way to write apps but it was pretty much a unique SDK per phone model. No app store of course.
It hit the market a few months after Apple revealed the iPhone 3G and the App Store.
When the Storm came out, the place I was working had placed pre-orders for the executive team. First day, they're all configured and sent out to much fanfare. Second day, many of them came back to IT, some airborne. After a month they were pretty much all returned in favor of something that actually worked, and we never placed any further orders.
Remember their 'haptic' touchscreen tech, SurePress? Click Clack. Felt like a child's toy. Quite the departure from genuinely great stuff like the trackball/pearl.
I remember the no Wi-fi. It was shocking, but made perfect sense from BB's perspective. These were work devices after all, and your precious data was expected to be routed through an encrypted BES server that would also compress packet contents to save bandwidth.
But it did have apps. Just not many. I remember downloading Pandora, Shazam and Whatsapp from the Blackberry App store. Android and iOS stores obviously dwarfed it in terms of quantity though.
Verizon was pretty hilarious with denial early on. I recall them coming to talk about what a piece of garbage the iPhone was to our CIO, and how BlackBerry would bury them, etc.
Then the CIO pulls out his iPhone and demos it. Lol. One of the reps flipped to AT&T a few weeks later.
It’s really a shame actually, the BB was a great device and had an unbeatable security model for enterprises for a long time. iPhone is just a magical product.
iphone is magical; yet I'm replying on pixel3 right now despite recently moving to 12 max pro. I'm now carrying 2 phones around. the iPhone for the camera and the pixel for the superior usability. magical I tell you.
At the time though, the iPhone was questionable for a business user on paper. Weird touch keyboard, poor security (its active sync client lied about encryption), etc.
That verizon would do that is not surprising. They did not want to budge on those dataplan costs. Touting 'superior network' as their selling point. It was not until they tried to move into M2M that they realized no one was going to pay 45 bucks per MB and that AT&T, T-Mobile, and Sprint would sell data and text at a much more reasonable rate. The original iPhone was neat but not much better than everything else out there in that category. Its one killer feature was that unlimited plan. That turned a remote device from something you did very little with because you did not want to get reamed on the data plans, to something you could hook the internet up to. It took verizon ages to realize it.
>Verizon pressured Blackberry to rush the release of their first touchscreen device
It's hilarious how higher-level managers always pretend that they do hyper-intelligent long-term planning, but in reality this is usually what they resort to when things go south. Just rush things. Make someone compromise engineering either via stupid features or working long hours or both.
Well, there must have been more to it than just "first touchscreen phones" - Openmoko Neo1973 was announced shortly before the iPhone was revealed and it was a touchscreen-only phone already, delivered in 2007.
> Sounds like they were just bogged down and crippled in a corporate way
They were tripped up by their own success as the 800 lb gorilla of dumb phones that sold in the hundreds of millions.
Nokia had a seperate department for “renewing the company through innovation”. It was an internal incubator which reported directly to the CEO. It could scout for tech, and develop prototype product products.
The problems started at the point where a new product was ready to leave the incubator. It had to be “sold” to an existing business unit in a suitable division.
The hurdle was that pre-2007 Nokia divisions were raking in billions selling mature tech in enormous volumes through existing channels.
There was no room and few candidates to champion anything new which required nurturing, development, ancillory technology and marketplaces, or, heaven forbid, new business models, starting from a revenue base of $0.00.
There was no shock when iPhone was launched. Nokia had tried touch display phones already by that time and they had gone the way of N900 and deemed unviable. Disarray yes, but not because of iPhone. Of course there were people that appreciated the threat that iPhone might become, but if anything it was laughed at and ridiculed. I mean come on - no 3G and no QWERTY keyboard. Who in the world would use that! ;)
I loved the Nokia N900's design, that hardware keyboard that made it so easy to do remote system administration over SSH, or keep doing everything that I want to do in Emacs just like on my laptop or desktop computers. Also like all Nokia phones of the time, it was nearly indestructible and could easily survive drops that would destroy a modern smartphone’s screen or casing.
However, in retrospect not all was rosy. Advocates of Linux phones today would prefer to avoid binary blobs that make upgrades impossible (the N900 is forever stuck on kernel 2.6) and a cellular modem connected directly to main memory.
I wish I had the opportunity to use the N900. I still have a partially working N9 (the SIM holder lock is broken meaning I have to hold the SIM in to use it for calls). I still browse and post on HN using it on occasion.
There's a spring mechanism that pushes it out. It worked by pressing in to lock it and then pressing in again to unlock it. The cover is recessed relative to the case, so it would be a bit difficult to secure it with tape unfortunately.
I still have my N9 lying around somewhere - had to get the sim tray replaced two or three times but otherwise I loved it. How did you update the security certificates? I tried to boot mine up a while back but it was basically unusable on the web because of HTTPS.
I haven't updated the security certificates so far, bit I don't think that most of the TLS connection issues are due to expired CA certificates in the certificate bundle. I believe it's because it can't negotiate a mutually agreed upon cipher.
It still works with a few websites like this one. In fact, I'm using my N9 to post this comment :-).
I had one of these and your note of "a cellular modem connected directly to main memory" makes me laugh. I didn't know that until now but I think that clears up why phone calls could crash the phone in some circumstances.
Like you said it was a fantastic little linux terminal - great for SSH and it was certainly fun and novel to have a linux CLI on my cell phone.
from my experience porting cyanogenMOD from android phones that only received 2.3 officially all the way up to 4+, you are likely running the binary kernel drivers as binary blobs for 2.6 and some kid in a forum either flipped bits at random while following a asm tutorial or wrapped it in a syscall shell that emulates an old kernel to the driver.
or maybe the n900 had less esoteric peripherals than android phones. I don't know... as I never got my hands on one despite actively trying to buy one for a couple years when they were announced.
Another PowerVR GPU for those that wonder. They quite notoriously don't have any open-source drivers.
There is a lot of information on the net why that's not the case, from the GPU architecture to source leaks. One could possibly make a shim to load older proprietary kernel modules, but it's hardly worth it: PowerVR GPUs don't seem to be used much nowadays, in contrast with Mali GPUs. And PowerVR GPUs seem to require quite different drivers depending on IP customization.
Maemo Fremantle may be stuck on 2.6, but N900 is not - it's pretty well supported in mainline kernel. Also, the cellular modem (Rapuyama - BB5) on N900 isn't connected to main memory, it's pretty much a separate component (which was used as a main CPU on some other Nokia phones).
I had a Nokia N800 "Internet Tablet" and loved that device. Android was a huge downgrade compared to what you could do with that in every way except interface, and the interface wasn't too bad.
I used the N800 until the battery eventually failed.
It was amazingly liberating, as a 'computer person', to have a truly hand-held computer; all my friends and family thought it was a stupid toy with a hard-to-understand interface -- but they liked drawing on it.
I had an N800 and found it so amazing, 2 sd card slots, pop out camera that could rotate, and a useful kickstand.
I recently found the Planet Computers devices and tried their first version out, it is conceptually a bit similar to the fxtec mentioned in the blog post.
I had the older 770, and loved it with all its limitations. I recall downloading Google maps images of a desired area trough a 3rd party tool that would also stitch them together, then and using them on the 770 with an external GPS receiver to have sort of a live map while driving. And the LCARS Star Trek styled screen gave a nice touch too.
Nokia was developing great products back then, then one day Microsoft came and ended it all.
Did the same with my N800 while driving in South Africa. The downloaded satellite maps showed the roads that were actually there, unlike those on the maps.. and that was very important, as it turned out. Very glad we had that N800.
(I had the N800 first, and got the N900 when it came out)
Same. In college I would show it off while working at the IT desk by sshing into the computer lab machines and checking on running programs/homework/etc. Used a bluetooth keyboard for the full "hacker-on-the-go" experience
My favorite device from that era was the HP Jornada since that with linux and an orinoco card felt like something straight out of the hacking movies. https://youtu.be/1s-tekLn7Nc
Thom makes a really insightful point: "people who want a true Linux computer in their pocket". I wanted a more open iPhone and I hated the N900 because it was terrible as a smartphone. Between the stylus, unconstrained multitasking, desktop Firefox, and Flash it's like they were trying to prove Steve Jobs right. The G2 was so much better because it had an actual phone OS.
I liked the stylus. More precise than a fingertip, which it wanted to be on a screen that small. Although you could use a fingernail too, in a pinch. People pointed at the resistive screen as evidence it wasn't as good as an iPhone, but the thing is that it was a very, very good resistive screen. High resolution, accurate, and you didn't need to mash it to get it to register.
On Android it's very constrained. Background apps are quickly killed to save battery, whereas on my N900 any running app which wouldn't stop by itself would drain it in a few minutes.
Apps can't even reliably sync data on Android without using a special Google-hosted remote service that pings your phone when there is an update. Music apps have to register specially so they don't get killed in the background.
Android is constrained in that sense, I agree. I thought the OP meant constrained like iPhones of the period were: No pretense of multitasking, every task switch being "close an app, open another one".
I have both an N900 and a Pro1 and it's everything I wanted in a phone/device. I'm using self-compiled LineageOS on it right now and it is as "Linux-y" as I need and want, complete with a usable terminal and SSH apps.
I'm not sure why you haven't been able to get a review unit of the new Pro1-x model, but I'll get in contact with someone at f(x)tec for you.
One thing from the N900 that I really miss on the Pro1 is the ability to multi-boot different operating systems. I think at one point I had five OS's running on my N900, mostly off of various microSD cards - Maemo, Android, Meego, SHR and Plasma Mobile. For the Pro1 I'd settle for being able to dual-boot Sailfish and LineageOS, but as it is now you need to pick one or the other.
Many people in the community are getting around this limitation by running virtual machines of the various operating systems they want. With the modern processor (compared to the N900) and the 6GB of RAM, performance seems to be totally usable on the Pro1.
When trying out various Linux handheld options before pmOS, I wasn't able to get N900 hardware at a good price, but I did get a few N810 units. Then I was dismayed to find that a lot of the once-public open source and tools for N810 had simply been removed from the Internet. That kind of thing is another reason to support efforts like pmOS.
> a lot of the once-public open source and tools for N810 had simply been removed from the Internet
Nokia insisted in hosting everything themselves, and obsessed with maintaining control of the platform while trying to "do opensource right" on some philosophical level. IIRC it resulted in a constant struggle with the community on really stupid issues, which dragged down the overall development speed. And then they pivoted to the Intel partnership just like that, from one day to the next, which meant jettisoning a massive amount of effort from the community (deb to rpm, effectively abandoning Qt, etc).
It was really a strange relationship. They kept writing rivers of (digital) ink on their love for OSS and blablabla, but then took all major decisions behind closed doors.
Interesting. I wasn't involved at the time, and, when I went trying to reconstruct the open source afterwards, I suspected the ill-fated Microsoft deal might've had something to do with it. But sounds like there were problems before then.
A lot of cross-organization open source projects have had trouble figuring out how to to do that. I think our collective understanding of this improved even within the last few years (e.g., community development processes like Rust's).
(Though it's not necessarily monotonic improvement: we might also be forgetting things that people used to know -- repeating mistakes that were already learned the hard way, and also making new mistakes that were easier for an earlier community foresee and avoid at the time.)
Another nice OS for the N900 is Maemo Leste (https://leste.meamo.org), which is basically the same software as ran originally, but modernized; sadly, when I last tried it, it still lacked a GUI for key features like calls and texts.
Maybe I wouldn't say "every way" (it doesn't age very well, several of my N900s have developed issues with their digitizers by now), but I do agree that it's superior. Much more accurate and pleasant to use than today's touchscreens; and way less prone to accidental touches. I have switched to a Librem 5 now, but I miss the N900's resistive screen.
The killer feature of a resistive screen IMO is its environmental resitance. It's not afraid of water, and can be used in gloves, or using any vaguely stick-shaped tool when your fingers are dirty.
Are you still stuck in 2008? Perhaps worked at Nokia?
I heard this a lot back then. Even believed it myself because how could 1000s of engineers be wrong... Ohh how nice those screens felt!
Then again it required only a few seconds of iphone usage to realize what a lie it is. Even the first crappy android phones with capacitive touchscreens was enough to show this.
> Then again it required only a few seconds of iphone usage to realize what a lie it is.
I'm going to have to call this subjective, because, surprisingly, I've also used capacitive screens.
I could use the N900 wet and in gloves, and it was pressure sensitive. Also, I once dropped it so hard that it broke a tiny piece of the sidewalk off (I tried to catch it while it was falling, and instead batted it and added velocity to the fall.) The screen was fine. I've never seen a cracked N900 screen.
The only time it required a stylus is when I was using a desktop UI on it, making everything microscopic. If you did that with a capacitive screen it would require a special stylus, whereas with my N900, a toothpick would do.
While definitely not "superior" in every way, it was more precise also with a finger (at least on the n900; other devices had worse quality resistive screens).
So you for example could draw or hit small buttons much more easily.
For capacitive screens you instead have to add another digitizer layer by wacom or n-trig to be able to draw, which took a while and still is not common enough.
That said, bad quality resistive screens were much, much more frustrating to use than bad capacitive screen.
I wouldn't say "every way", but the lack of multi-touch was sometimes an advantage. I loved the swirl-to-zoom gesture on the N900, pinch-to-zoom on multi-touch screens always feels clunky in comparison. If I wanted to zoom in or out on the N900, I just used the thumb that was already hovering over the screen, of the hand holding a phone. Pinching to zoom needs two hands, one to hold the phone, the other to perform the gesture.
Amen. I absolutely hate pinch to zoom. Screw to zoom is better, and actually educational. Teaches righty-tighty lefty-loosey.
IIRC lack of multitouch was a software problem. The N900 was rife with software problems. If they had just kept the Linux orthodox and left out the binary blobs, they would have been able to go back to the platform after the MS disaster with the OS looking and working 10x better than where they left it.
This was a neat phone to have during it's time, but I remember having it for a while and the killer app that came out months after I got the phone was MMS. The engineers at Nokia just couldn't figure it out but someone hacked on it long enough to enable it. Fairly certain they flew the guy out and offered him a job.
Nokia had actually done the groundwork for a fair bit of MMS support and they were really helpful with answering questions (they even shipped a list of access point names in PR1.2 to make fMMS much easier to set up)... But yepp, I ended up working at Nokia afterwards :D
I remember building simple PyQT apps on the N900, and then C++/Qt. At one point I was very tempted to start my own business and become a real developer for the platform while it was still "young", but it became apparent very quickly that Nokia was not seriously invested in it, and in any case they were not interested in small devs (getting an account for their joke of a store was a byzantine process, and expensive too iirc).
I attended an event with an "evangelist" here in Manchester, I was all fired up. It was some dude from Finland on a UK tour, and from the start it was all about the transition to QT on Symbian - this was to an audience full of people already working on iOS, to whom Symbian was a bad joke. He kept going on about the marvels of QT - which at that point was already a decade old and hardly a novelty, but clearly it had just reached Finland or something. Maemo was a footnote, "oh yeah we have that too", it was absolutely clear he just didn't care for it - the future was QT on Symbian. People were snickering; I wanted to strangle him. I reckon he single-handedly destroyed any local interest in developing for Nokia. I gave up shortly afterwards, and then the burning-platform memo happened and that was it.
My N900 is now an mp3 player for my 9yo son. The MicroUSB-B socket is a bit loose, iirc that was common with early adopters of that standard. The backlit keys have yellowed, or maybe the LEDs have. Everything else still works fine, although the software is obviously obsolete. My kid wonders why he has to press the screen that hard, and I feel like someone trying to explain ancient history.
> He kept going on about the marvels of QT - which at that point was already a decade old and hardly a novelty
From what I remember this was because they'd spent so much time and effort building up the maemo (or meeGo or whatever) and burned up so much goodwill with the switch to Qt. When the burning platform memo came out they'd literally just burned another platform, regardless of technical merit the timing for moving to Qt couldn't have been worse.
Big +1 to this. I was a big n900 fan who ended up getting a job a Nokia through some of the OSS stuff I made for it and your diagnosis of the “Qt for Symbian” problem is 100% spot on. The amount of time/effort that they put on Symbian past 2009 is ridiculous... classic case of the fallacy of sunken cost.
I was also pretty enthusiastic about the platform. Nokia big missed opportunity. My impression was that there was a brief window in history where they could've set themselves up as a rival to iOS and Android.
If the Nokia board had not hired Elop or had fired him in time, Maemo would have been a third mobile platform alternative today - one likely far more open than Android.
Elop was a truly successful Nokia saboteur. An illustrative case study of how a single CEO can destroy a company for his self-profit and go out laughing with no legal repercussions.
> A Finnish newspaper has uncovered information in the SEC filings for Nokia's sale to Microsoft that show former Nokia CEO Stephen Elop had a huge incentive to unload the company in the form of a $25.5 million US bonus that he would get in the event of a "change of control" in the mobile company.
Yes, please do your own research and read on how the Nokia's board of Directors saw the memo as an act of misjudgement and how Chairman Jorma Ollila gave bitter feedback for it at a board meeting.
During Elop's tenure, Nokia's stock price dropped 62%, their mobile phone market share was halved, their smartphone market share fell from 33% to 3%, and the company suffered a cumulative €4.9 billion loss
From the inside that is how pretty much we went through it.
The community was already getting tired of the multiple Symbian reboots (just IDEs there were three of them), then came PIPS and Qt, while it seemed like a strecht everyone was kind of still on boat.
However then the whole Linux vs Symbian started to gain steam, Symbian went open source, and finally the burning platforms memo happenend.
Not many happy faces at Espoo during that week, and even more unhappy devs asked to throw everything away and jump into .NET, when the Nokia community had a long tradition of Java and C++ development stacks.
In Finland, I think tech people thought Symbian was bad too, and Maemo and Meego or web applications on mobile etc were the future. Finland has a very strong open source culture among the doers, most of the education at Aalto university uses open source software etc..
Unfortunately Nokia was not able to transform to a software platform company, from what I heard, management didn't really understand anything of that.
It's kind of funny that nothing notable in the open source world came out of Aalto.
Linus Torvalds and Tatu Ylönen (ssh) studied in University of Helsinki while University of Oulu's Jarkko Oikarinen created IRC.
Yes, sorry, last night I was a bit flippant and this issue remains a bit of a thorn in my geek heart. I was trying to convey how this guy, flown out by Nokia HQ, felt and sounded like coming from a parallel dimension where developers really wanted to use Symbian in 2008/2009.
I apologize for any offense I might have caused to Torvalds' homeland.
Symbian folks inside Nokia at the time lived in a reality distortion field. Your experience is not the only instance where the Symbian division inside the company railroaded the Maemo project, which they saw first and foremost as a threat to their jobs and political dominance. As TFA notes, N900 was preceded by N770, N800, and N810 "internet tablets". N810 in particular was basically the same device as N900, but despite having otherwise state of the art wireless communication tech all of them lacked one very crucial component: a mobile data modem. Internet connection was only possible over WiFi (or even RJ45-USB dongle) but not SIM.
The reason for this was of course not technical - the Symbian folks inside Nokia simply managed to convince the executive level management that adding mobile data would turn these devices from PDAs to a smartphones, and it would be very ill-advisable to launch a competing smartphone platform to Symbian.
When Maemo folks finally got a chance to "give us a try" (and I guess management saw with the competition that they needed something more modern), the result was N900. Considering it was meant as a "niche" device its sales exceeded all projections. However, in the grand scheme of things it was just an experiment designed to fail and further cement the dominance of Symbian.
I think Nokia also developed a fully-working tablet around either Maemo or MeeGo. It was literally ready to ship, but was at last minute cancelled before the launch event due to backroom dealing again by the Symbian folks.
It has all been documented in the book "Operation Elop", which paints a pretty good picture about everything that went on behind the scenes during the demise of the once great Nokia: https://asokan.org/operation-elop/ The book is based around absolutely exclusive interviews of people who served in Nokia's board and top executive positions at the time.
Maemo was the right platform at the right moment to compete with Android and iOS and I was very surprised they closed it down in the time of crisis instead of realising they had built exactly what was needed.
It had a modern OS with memory overcommit instead of the endless drudgery of checking out-of-memory conditions as in Symbian.
The N770 launched with a GUI based on Gtk and convincing OSS developers to move to Qt in later versions of the platform was bump in the road.
(I think Gtk had problems with basing the inheritance model on matching strings instead of vtables as with C++, so performance on this type of devices was problematic. Or - someone please correct me. I know it feels a bit weird given the long history of Gtk.)
It wasn't about performance; GTK (or rather a GTK-based framework called Hildon) continued to be the default toolkit all the way to Maemo 5 ("Fremantle"), which is what shipped on the N900. It was also what Intel used on Moblin, the OS that was supposed to merge with Maemo to become MeeGo.
The issue with QT was all about Nokia: QT was supposed to save the Symbian platform, and also provide an on-ramp to Maemo (at least in theory - in practice nobody was really committed to that...). So the Maemo folks were obligated to switch to QT, which they did fairly easily; the result was what shipped on the N9, effectively Maemo 6 ("Harmattan") rebranded as "MeeGo 1.2".
In short: QT was adopted to please the Symbian people, and it was all for nothing anyway. That suited me just fine when first announced (Hildon was an under-documented mess of C with obsolete and unusable Python bindings, whereas QT had first-class wrappers like PyQt), but many Maemo old-timers never really warmed to it.
I used my N900 until a week ago, I finally moved the SIM card over to another phone (a dual-SIM one), because the (original) battery didn't take much charge anymore. Hardware-wise it's as good as new (mine is Made in Finland). Over the years I've used it to do support work in Antarctica using VPN from whatever cafeteria I was at at the time, other types of work where I needed a Linux computer and only had the phone, it's been running my minicomputer emulator, I did development work for it (the SDK running on my PC - basically a Debian setup), and, of course, as my phone.
The MicroUSB socket was a weak point, the fix was to file off the two notches on microUSB cables before using them with the N900. So my N900's USB socket is still good as new.
I cut my teeth on GUI development with PyQT back in the day. I couldn't afford an N900 at the time, but was sorely tempted.
Ironically the best mobile platform for hacking GUI and mobile apps in Python now is, er, iOS. Pythonista gives access to almost the entire iOS API suite (and slightly hacky access to all of it), even has a GUI builder built-in, and is highly programmable directly on the device. Pyto is pretty cool too. Hard to believe, but that's where we are.
You still cannot ship the resulting apps though. One has to hack around with the likes of Kivy to get that. On Maemo you could just ship a Deb (at least outside the braindead Nokia store, which was restricted to C++).
The sad thing is, I remember attending a PyConUK about 10 years ago where this was a clear item in the keynote (by Van Lindberg, iirc): "the Python story on mobile is non-existent". Everybody agreed it was a priority.
Several apps have been shipped based on Pythonista, including games. The developer released an Xcode template, you drop your Python code into it and build an app from it. Pyto is open source, once you have a working app developed on-device you can just pull Pyto from Github and add your code to build an app the same way, but it's not as complete as Pythonista.
There are non-python dev environments that enable the same thing, like Codea that uses Lua. It has several published apps and games out there on the App Store as well.
Python on iOS is still a joke compared to what you can do on phones like Librem 5 or PinePhone, or even several other devices via distros like postmarketOS.
In what way? On my devices I have full access to the native APIs and robust networking libraries for access to web services. There's Git integration available. I'm not aware of any Android programming environments that provide on-device GUI builders. I can sync code using iCloud or Dropbox if Git is too heavyweight.
Frankly the options on Android look extremely primitive in comparison, and largely depend on off-device development for anything above very basic and incomplete terminal prompt level features. iOS on-device development went beyond that with commercial-grade on-device development environments a decade ago, but Android seems stuck there permanently.
I'm not really sure why that is, there doesn't seem to be any technical limitation preventing the emergence of well specified complete development environments, like Codea, Pythonista and Pyto on Android but it never seems to happen. When I got my first iPad I was half expecting to have to get an Android device eventually just to be able to code on it, but it never happened and the on-device development story on iOS has gone from strength to strength.
In a way that Python is provided by the system out-of-box, you can use regular PyGObject bindings for everything (or Pyside or whatever else you want), use any third party modules from system repos or pip - exactly how you would on a desktop; you can even use it to write system daemons. Neither iOS or Android can compete with that.
"and in any case they were not interested in small devs (getting an account for their joke of a store was a byzantine process, and expensive too iirc)"
I had the exact same experience with RIM/Blackberry 957-era... Great device, completely clueless company regarding an independent software ecosystem.
Say what you will about Apple, but they truly knocked down walls and opened the mobile floodgates to independent developers.
I still hope some viable and fully open Linux option will emerge that's not Android. And have good quality devices with that. Google did a major disservice to Linux at large by the rift it caused.
I don't want bionic with Surface Flinger. Give me a Wayland compositor and a normal Linux stack.
That's a very good development for sure. I hope the hardware will be able to progress, since SoC choices for further improvements are very limited and the likes of Qualcomm are DOA.
I was so excited about the N900 when it was announced - a linux machine in my pocket! I couldn't justify the price, being freshly out of school though. Nokia advertised a competition asking the public how they would "hack" the device to make something cool. I submitted a proposal to make a bike dashboard - it sounds so basic now, but I wanted a simple way to track my ride and proposed fun little additions like using the ambient light sensor to turn on an external headlight, a horn, etc.
They picked my proposal along with a couple others and gave us some budget to buy hardware and components, and about 3 or 4 weeks to complete the project. It was my first time soldering, first time using an Arduino, and I remember being so stressed out about whether or not it would work.
It all came together in the end and looking back with hindsight now, I really owe a lot to that competition and the phone that let me pretty quickly write python apps on it :) Good times.
I got it as a graduation gift and it steered me into this field due to giving me a real contact with Unix/Linux and tinkering. I had tried Linux desktop but as a young gamer the moment browsing the web for the day was done, it was time to fire up Windows and play.
The N900 was different. I didn't understand everything I was doing, but installing deb's, overclocking a CPU, installing Android v2.x, installing emulators, browsing forums for more tinkering (I could go on) was the thing that made me love computers more.
Last but not least, it had a killer feature any fresh 18-year-old with friends that had cars loved back then: FM transmitter!! Back when AUX inputs were still a rarity and bluetooth was pairing... pairing... pairing... just pressing a button and tuning to a channel to get crystal clear audio was AMAZING. That and the IR blaster made for some nice party tricks.
For the best price/performance ratio, I will still go with XA2 for now. Here, I can get a used XA2 Ultra in like-new condition for less than $200. Sailfish 4.0 is fantastic, and with micro-g, the phone works perfectly fine as a daily driver.
The same couldn't be said for N900 or N9, at least for me. J1 was rather slow but good enough, Xperia X was great for a year or so until Android 4.4 got churned, and now with Xperia XA2 and Xperia 10, we are finally there.
Heh. The N900 was an amazing device and I cherish the time I had with it.
However, it's also a painful memory, because I managed to brick the device after spending possibly hundreds of hours customizing it.
The first thing I did was let the micro USB port come loose--a common problem with the device, which seemed to happen to many over time. I remember still being able to charge it somehow (maybe I bought an external charger?). However, the nail in the coffin was when I was cleaning up some "unnecessary" cruft in the init directory, ostensibly to make the thing boot faster.
What I discovered instead was that there was a watchdog that monitored for the existence of crucial files, and, failing to find them, the phone would reboot itself. So my glorious N900 was caught in a bootloop, and moreover without a functioning USB port to re-flash the operating system.
Behind the battery, there are CPU pins that are directly exposed through a grid of copper pads. I made an attempt at wiring up a makeshift soldering job directly to the pins corresponding to a USB port, but failed to confidently pull off the dexterity to work with the small surface, as well as the care to do so without heating things up too much.
Of course I believe there existed a development device with pins that directly clamped onto the pads, but if such a thing was publicly available it was probably hard to come by at affordable prices.
I later tried to realize a similar experience to the N900 on Android using stuff like tmux and F-Droid. However, I have no good memories of any phone after my Nokia, and pretty much don't like phones at all at this point, and now just use my iPhone SE as a glorified flip phone with maps and a web browser.
Mine died to this. USB port came loose, sat in the back of a drawer waiting for me to get round to fixing it, to be thrown away in a fit of "I'm never going to get round to any of these 20+ projects-in-waiting that have been collecting dust for the past N years". The battery was probably knackered by that point anyway.
I still have one of these someplace in a drawer. Maemo was pretty nice, but the UX was all over the place sometimes and it became tedious to click through some menus with the stylus.
Running a browser on it was doable, but the bit that really became a blocker for me was the lack of a fast, simple e-mail client (even non-touch screen Blackberries were better, and had slightly better keyboards).
I still have one in a drawer. I worked at Nokia at the time. These were awesome phones. But they were deliberately crippled by management that preferred other platforms they had.
So, several devices originally earmarked for Maemo/Meego (the name changed at some point) actually shipped with Symbian instead.
Some fun facts about Maemo:
Shipped in 2006 with the N770. Had a tablet UI based on X and GTK, a mozilla based browser and was based on Debian. You could build your own packages, add your own apt repositories, and people basically built most of Debian from source to run on it. E.g. I had a JVM running on it and a full LAMP stack at some point.
The N800 was for a time the only device that could run Android (dual boot) and basically shared large parts of the kernel with it. Google used this device for early Android testing and probably owes a lot to Nokia engineers contributing patches to the Linux kernel. The first Nexus phone shipped quite late in the Android development process.
Development of Meego saw several strategy changes that ultimately delayed products for a long time. One of those unfortunate decisions was the decision to swap out GTK for QT; effectively retiring all of the user facing UI. That set back the clock by years. And it got slowed down further by a simultaneous attempt to get Symbian on QT as well. That trainwreck of course ended up sucking up all the resources and failed to ultimately deliver the goods. This was in the middle of Nokia's realization that "oh fuck this iphone thing is real; we need a touch screen UI". It killed at least two touch screen platforms that it had before it came to that realization.
Technically, the N800 was running circles around the first iphone in terms of what you could do on it. All it needed was a bit of hardware polish and a phone stack. Early Android had nothing on Meego. Even the UX was clumsy and it was slow and limited as well. But unlike Meego, Google shipped it and supported it and persisted in developing it. Nokia instead threw out the baby with the bathwater betting on several other horses before ultimately walking away from the whole phone business.
Nokia declined to combine a phone stack + SIM card with the N770 and N800 deeming it to risky; they were protecting their deals with operators to basically cripple phones in favor of crappy operating services. E.g. bundling Skype with the n800 was controversial. It even did video calls with Skype. In 2007. The N800 was a tablet before the iphone, ipad or Android were a thing. There was not a lot else in the market at that point. That thing with a phone stack would have been a killer product despite its many limitations.
The N900 and the later N9 were both labelled as developer phones and cut off from any serious marketing effort. The N9 technically was a decent phone but by the time it shipped, the team had been layed off, the platform cancelled and Nokia was endorsing Windows Phone. It also shipped an Android phone around the same time, leaving the ugly job of cancelling that product to Microsoft after the acquisition.
Samsung's Bada is a direct decendent of Meego. It never really gained any market share against Android and it seems Samsung pretty much gave up on it.
I would be curious to know more about the several devices that were planned to be on Meego but then shipped with Symbian!
I remember seeing Bada phones at the time, but (at least in my country) the "does it run WhatsApp" era was already starting, and the total lack of apps for Bada made it totally unattractive.
All said, I am not sure that polishing and marketing would have been enough to save the situation. I didn't see Palm's Pre and webOS mentioned in the thread so far: that makes for an example of an excellent platform, with good marketing and capabilities, that still didn't manage to affect the rise of the duopoly.
I know the N8 was at some point earmarked for Meego. But internal politics happened.
That would have been a true flagship phone around 2011 with a 12 megapixel camera, aluminium body and nice touch screen. I had it in Symbian form and it was alright but ultimately meh in terms of software. Essentially all Android phones at that point were a combination of slow, not very premium, and riddled with bugs. This was way before Google figured out how to do software updates properly.
A slick UI, linux, touchscreen, etc. exactly what Nokia needed at that point. But they convinced themselves Symbian was good enough and were corrected by the market that made it very clear that it wasn't even close. Of course crippling the device with not enough memory did not help. Nokia saved pennies and sacrificed market share with that.
It's stunning to me just how many incredible technical efforts Nokia made to just kill.
The N9 after this was killer but boom, dead, the Nokia X was okay but weird, Dead, the Lumia, alive, then boom, dead.
It's remarkable that Elop was never prosecuted for his actions that clearly were in benefit to his prior employer.
It's almost hilarious that modern Nokia HMD makes phones now because I expect the company to randomly up and disappear and come back with a WebOS phone or some other weird offshoot.
Elop was a stooge that executed the master plan created bv the Nokia board together with MS as specified. Blame the decade of incompetence that preceded Elop.
> One of those unfortunate decisions was the decision to swap out GTK for QT;
That was kind of swings and roundabouts, though. The second version of the Maemo app I worked on ended up being a major rewrite Gtk->Qt thanks to the switch, but Qt was so much better.
Yes, it just meant shipping after Apple and Android had become entrenched in the market. Nokia wasted a good two years doing this. And it was too little too late.
I was a Symbian developer at that time. The fate of N900 is a grim reminder that engineering brilliance will still lose out to managerial incompetence.
Seriously, the device was just so attractive both for users and programmers. If Nokia were able to stay the course, they could pull off a bestseller.
Their nosedive into the ground under Stephen Elop is a stuff of dark legends.
Not really. Even before Elop, there were multiple layer of management and the culture of CMA (Cover My Arse) was so strong, as I heard from other Nokia ex-employees. Not to mention people from Symbian Inc wasn't good people according to the book about rise and fall of Symbian.
It's interesting how devices have gone.
Nokia, when they were focused, were making devices smaller, cheaper and with improved battery life. Now we get these giant devices that need re-charging daily and cost >£1000
It's a good point. I think it highlights the fact that these are not really the same kinds of device, so can't be evaluated on the same criteria. If we were still just buying phones nobody would spend that much on something that consumes so much power. The mistake Nokia and the others made was they thought the iPhone was just a phone.
181 comments
[ 3.4 ms ] story [ 238 ms ] threadWhen this was out I was using the OG Motorola Droid. I wanted one of these but they were not available on my carrier so I stayed on the Droid line. The hot-dog slide keyboard was amazing. I had full terminal control from these.
Nowadays I actually still rock a 2017 iphone SE; dimensions wise I don't want anything bigger in width or height, but i would gladly add some thickness to get a full tactile keyboard again.
Lately the "main" button has been failing, and since tapping the screen doesn't get it out of sleep anymore, I can't pick up calls anymore. So to access the phone, I get out the SIM and re-insert it.
It's time to abandon it here too, but it was an excellent phone in use for almost 10y.
> since tapping the screen doesn't get it out of sleep anymore, I can't pick up calls anymore.
Mine will show the the screen and prompt me to swipe up to answer the call. Does yours work in a different way, or is the touch system failing?
I'm now more thinking about continuing that using AndroWish [2] instead (Tcl/Tk on Android).
My N9 in better times did exactly that when someone called, but now the screen stays black and nothing reacts (it still rings though). I don't know what causes this, because out of the sleep mode then touchscreen works fine. I have to pull out the SIM everytime as such, put it back in and log back into the network (and then often call back myself ofcourse ;)). Not going to bother anymore, but I suspect a software problem. It also started to crash/hang it's applications from time to time.
[0] https://en.wikipedia.org/wiki/MeeGo#Mer
[1] https://www.equi4.com/tclkit/
[2] https://www.androwish.org/
I think we have Stephan Elop to thank for that. I never really understood why Nokia didn't just focus on the market outside the US. Had they continued producing phones like the N9, I'm sure they could have retained a lot of market share in Europe, Africa, Middle East and Asia.
> I would still be using it if the CAs hadn't expired
I'm not sure whether all of the CAs had expired, but I wonder if it would be possible to get an updated package with up to date CAs. The main issue I have is that it can't negotiate an agreed upon cipher when trying to establish a TLS connection.
I'm still able to use the Firefox browser to browse a number of websites like Facebook and Reddit. Hacker news still works with the built-in browser. Facebook sometimes works with the default browser and the old subdomain reddit website will lock up the default browser in my experience (the current reddit website layout is a no-go).
Apparently there was an ongoing joke inside of the company that every button on these phones was one of the executives. And so you had multiple buttons/ways of doing one thing because consensus. Sounds like they were just bogged down and crippled in a corporate way that so many European companies were between 1990-2020.
Apparently when the iPhone was announced the whole company was in deep shock and disarray. They were planning on the first touchscreen phones hitting the market around 2010-2015, and here was a company promising to deliver one by 2008.
Because the iPhone was an AT&T exclusive in the US, Verizon pressured Blackberry to rush the release of their first touchscreen device, the BB Storm. It was released as a Verizon exclusive in late 2008, and coincided with the release of HTC/Tmobile G1, the first Android phone.
BB tried to argue that BBOS6 was a click trackball based OS and would not transition to touch well. In the end they went for a bizarre hybrid design where the screen pushed down when you tapped something on the screen. Like a button would. I got used to it over time but the first time I used it, I thought I'd dislocated the screen.
Within a year, Verizon realized it wasn't going to rival AT&T w/ BB, and had Motorola create the Droid brand, which may have been one of the most popular Android phones in the US before Samsung arrived.
Webpage would render incredibly slowly because Blackberry weren't capable of putting a wifi chip in the handset (it was limited to edge if I recall). The whole UI was sluggish. Cherry on top? No apps. There was a way to write apps but it was pretty much a unique SDK per phone model. No app store of course.
It hit the market a few months after Apple revealed the iPhone 3G and the App Store.
Remember their 'haptic' touchscreen tech, SurePress? Click Clack. Felt like a child's toy. Quite the departure from genuinely great stuff like the trackball/pearl.
But it did have apps. Just not many. I remember downloading Pandora, Shazam and Whatsapp from the Blackberry App store. Android and iOS stores obviously dwarfed it in terms of quantity though.
Then the CIO pulls out his iPhone and demos it. Lol. One of the reps flipped to AT&T a few weeks later.
It’s really a shame actually, the BB was a great device and had an unbeatable security model for enterprises for a long time. iPhone is just a magical product.
At the time though, the iPhone was questionable for a business user on paper. Weird touch keyboard, poor security (its active sync client lied about encryption), etc.
But in the flesh, it was magical.
It's hilarious how higher-level managers always pretend that they do hyper-intelligent long-term planning, but in reality this is usually what they resort to when things go south. Just rush things. Make someone compromise engineering either via stupid features or working long hours or both.
They were tripped up by their own success as the 800 lb gorilla of dumb phones that sold in the hundreds of millions.
Nokia had a seperate department for “renewing the company through innovation”. It was an internal incubator which reported directly to the CEO. It could scout for tech, and develop prototype product products.
The problems started at the point where a new product was ready to leave the incubator. It had to be “sold” to an existing business unit in a suitable division.
The hurdle was that pre-2007 Nokia divisions were raking in billions selling mature tech in enormous volumes through existing channels.
There was no room and few candidates to champion anything new which required nurturing, development, ancillory technology and marketplaces, or, heaven forbid, new business models, starting from a revenue base of $0.00.
However, in retrospect not all was rosy. Advocates of Linux phones today would prefer to avoid binary blobs that make upgrades impossible (the N900 is forever stuck on kernel 2.6) and a cellular modem connected directly to main memory.
Why not just tape it in place?
It still works with a few websites like this one. In fact, I'm using my N9 to post this comment :-).
Like you said it was a fantastic little linux terminal - great for SSH and it was certainly fun and novel to have a linux CLI on my cell phone.
That's good, because that isn't true. Rapuyama has its own 128MB of DRAM on the N900.
or maybe the n900 had less esoteric peripherals than android phones. I don't know... as I never got my hands on one despite actively trying to buy one for a couple years when they were announced.
There is a lot of information on the net why that's not the case, from the GPU architecture to source leaks. One could possibly make a shim to load older proprietary kernel modules, but it's hardly worth it: PowerVR GPUs don't seem to be used much nowadays, in contrast with Mali GPUs. And PowerVR GPUs seem to require quite different drivers depending on IP customization.
That has been done already: https://github.com/openpvrsgx-devgroup
https://www.fxtec.com/pro1
I used the N800 until the battery eventually failed.
It was amazingly liberating, as a 'computer person', to have a truly hand-held computer; all my friends and family thought it was a stupid toy with a hard-to-understand interface -- but they liked drawing on it.
(I had the N800 first, and got the N900 when it came out)
I'm not sure why you haven't been able to get a review unit of the new Pro1-x model, but I'll get in contact with someone at f(x)tec for you.
https://wiki.postmarketos.org/wiki/Nokia_N900_(nokia-n900)
There was also a project to replace the hardware:
https://neo900.org/
When trying out various Linux handheld options before pmOS, I wasn't able to get N900 hardware at a good price, but I did get a few N810 units. Then I was dismayed to find that a lot of the once-public open source and tools for N810 had simply been removed from the Internet. That kind of thing is another reason to support efforts like pmOS.
Nokia insisted in hosting everything themselves, and obsessed with maintaining control of the platform while trying to "do opensource right" on some philosophical level. IIRC it resulted in a constant struggle with the community on really stupid issues, which dragged down the overall development speed. And then they pivoted to the Intel partnership just like that, from one day to the next, which meant jettisoning a massive amount of effort from the community (deb to rpm, effectively abandoning Qt, etc).
It was really a strange relationship. They kept writing rivers of (digital) ink on their love for OSS and blablabla, but then took all major decisions behind closed doors.
A lot of cross-organization open source projects have had trouble figuring out how to to do that. I think our collective understanding of this improved even within the last few years (e.g., community development processes like Rust's).
(Though it's not necessarily monotonic improvement: we might also be forgetting things that people used to know -- repeating mistakes that were already learned the hard way, and also making new mistakes that were easier for an earlier community foresee and avoid at the time.)
It does not require a stylus, and is superior to capacitive touchscreens in every way.
I heard this a lot back then. Even believed it myself because how could 1000s of engineers be wrong... Ohh how nice those screens felt!
Then again it required only a few seconds of iphone usage to realize what a lie it is. Even the first crappy android phones with capacitive touchscreens was enough to show this.
I'm going to have to call this subjective, because, surprisingly, I've also used capacitive screens.
I could use the N900 wet and in gloves, and it was pressure sensitive. Also, I once dropped it so hard that it broke a tiny piece of the sidewalk off (I tried to catch it while it was falling, and instead batted it and added velocity to the fall.) The screen was fine. I've never seen a cracked N900 screen.
The only time it required a stylus is when I was using a desktop UI on it, making everything microscopic. If you did that with a capacitive screen it would require a special stylus, whereas with my N900, a toothpick would do.
For capacitive screens you instead have to add another digitizer layer by wacom or n-trig to be able to draw, which took a while and still is not common enough. That said, bad quality resistive screens were much, much more frustrating to use than bad capacitive screen.
IIRC lack of multitouch was a software problem. The N900 was rife with software problems. If they had just kept the Linux orthodox and left out the binary blobs, they would have been able to go back to the platform after the MS disaster with the OS looking and working 10x better than where they left it.
Nokia had actually done the groundwork for a fair bit of MMS support and they were really helpful with answering questions (they even shipped a list of access point names in PR1.2 to make fMMS much easier to set up)... But yepp, I ended up working at Nokia afterwards :D
I attended an event with an "evangelist" here in Manchester, I was all fired up. It was some dude from Finland on a UK tour, and from the start it was all about the transition to QT on Symbian - this was to an audience full of people already working on iOS, to whom Symbian was a bad joke. He kept going on about the marvels of QT - which at that point was already a decade old and hardly a novelty, but clearly it had just reached Finland or something. Maemo was a footnote, "oh yeah we have that too", it was absolutely clear he just didn't care for it - the future was QT on Symbian. People were snickering; I wanted to strangle him. I reckon he single-handedly destroyed any local interest in developing for Nokia. I gave up shortly afterwards, and then the burning-platform memo happened and that was it.
My N900 is now an mp3 player for my 9yo son. The MicroUSB-B socket is a bit loose, iirc that was common with early adopters of that standard. The backlit keys have yellowed, or maybe the LEDs have. Everything else still works fine, although the software is obviously obsolete. My kid wonders why he has to press the screen that hard, and I feel like someone trying to explain ancient history.
What a waste, Nokia, what a waste.
From what I remember this was because they'd spent so much time and effort building up the maemo (or meeGo or whatever) and burned up so much goodwill with the switch to Qt. When the burning platform memo came out they'd literally just burned another platform, regardless of technical merit the timing for moving to Qt couldn't have been worse.
Elop was a truly successful Nokia saboteur. An illustrative case study of how a single CEO can destroy a company for his self-profit and go out laughing with no legal repercussions.
Elop did exactly what the Nokia board asked him to do.
All anti-Elop bashers should read about what happened before spreading this content.
> A Finnish newspaper has uncovered information in the SEC filings for Nokia's sale to Microsoft that show former Nokia CEO Stephen Elop had a huge incentive to unload the company in the form of a $25.5 million US bonus that he would get in the event of a "change of control" in the mobile company.
https://www.cbc.ca/news/business/stephen-elop-to-get-25-5m-f...
Followed by Tomi T Ahonen posts,
https://communities-dominate.blogs.com/brands/nokia/
During Elop's tenure, Nokia's stock price dropped 62%, their mobile phone market share was halved, their smartphone market share fell from 33% to 3%, and the company suffered a cumulative €4.9 billion loss
I happened to be in Espoo office during the week the memo came out, to give a training on NetAct related technologies.
No research needed.
If they had set full sail on Maemo, it would have been the 3rd platform today.
The community was already getting tired of the multiple Symbian reboots (just IDEs there were three of them), then came PIPS and Qt, while it seemed like a strecht everyone was kind of still on boat.
However then the whole Linux vs Symbian started to gain steam, Symbian went open source, and finally the burning platforms memo happenend.
Not many happy faces at Espoo during that week, and even more unhappy devs asked to throw everything away and jump into .NET, when the Nokia community had a long tradition of Java and C++ development stacks.
Unfortunately Nokia was not able to transform to a software platform company, from what I heard, management didn't really understand anything of that.
I apologize for any offense I might have caused to Torvalds' homeland.
The reason for this was of course not technical - the Symbian folks inside Nokia simply managed to convince the executive level management that adding mobile data would turn these devices from PDAs to a smartphones, and it would be very ill-advisable to launch a competing smartphone platform to Symbian.
When Maemo folks finally got a chance to "give us a try" (and I guess management saw with the competition that they needed something more modern), the result was N900. Considering it was meant as a "niche" device its sales exceeded all projections. However, in the grand scheme of things it was just an experiment designed to fail and further cement the dominance of Symbian.
I think Nokia also developed a fully-working tablet around either Maemo or MeeGo. It was literally ready to ship, but was at last minute cancelled before the launch event due to backroom dealing again by the Symbian folks.
It has all been documented in the book "Operation Elop", which paints a pretty good picture about everything that went on behind the scenes during the demise of the once great Nokia: https://asokan.org/operation-elop/ The book is based around absolutely exclusive interviews of people who served in Nokia's board and top executive positions at the time.
It had a modern OS with memory overcommit instead of the endless drudgery of checking out-of-memory conditions as in Symbian.
The N770 launched with a GUI based on Gtk and convincing OSS developers to move to Qt in later versions of the platform was bump in the road.
(I think Gtk had problems with basing the inheritance model on matching strings instead of vtables as with C++, so performance on this type of devices was problematic. Or - someone please correct me. I know it feels a bit weird given the long history of Gtk.)
The issue with QT was all about Nokia: QT was supposed to save the Symbian platform, and also provide an on-ramp to Maemo (at least in theory - in practice nobody was really committed to that...). So the Maemo folks were obligated to switch to QT, which they did fairly easily; the result was what shipped on the N9, effectively Maemo 6 ("Harmattan") rebranded as "MeeGo 1.2".
In short: QT was adopted to please the Symbian people, and it was all for nothing anyway. That suited me just fine when first announced (Hildon was an under-documented mess of C with obsolete and unusable Python bindings, whereas QT had first-class wrappers like PyQt), but many Maemo old-timers never really warmed to it.
The MicroUSB socket was a weak point, the fix was to file off the two notches on microUSB cables before using them with the N900. So my N900's USB socket is still good as new.
That should be a whole post by itself.
Ironically the best mobile platform for hacking GUI and mobile apps in Python now is, er, iOS. Pythonista gives access to almost the entire iOS API suite (and slightly hacky access to all of it), even has a GUI builder built-in, and is highly programmable directly on the device. Pyto is pretty cool too. Hard to believe, but that's where we are.
The sad thing is, I remember attending a PyConUK about 10 years ago where this was a clear item in the keynote (by Van Lindberg, iirc): "the Python story on mobile is non-existent". Everybody agreed it was a priority.
There are non-python dev environments that enable the same thing, like Codea that uses Lua. It has several published apps and games out there on the App Store as well.
It was completely off my radar, clearly - it could use more publicity then...
Frankly the options on Android look extremely primitive in comparison, and largely depend on off-device development for anything above very basic and incomplete terminal prompt level features. iOS on-device development went beyond that with commercial-grade on-device development environments a decade ago, but Android seems stuck there permanently.
I'm not really sure why that is, there doesn't seem to be any technical limitation preventing the emergence of well specified complete development environments, like Codea, Pythonista and Pyto on Android but it never seems to happen. When I got my first iPad I was half expecting to have to get an Android device eventually just to be able to code on it, but it never happened and the on-device development story on iOS has gone from strength to strength.
I had the exact same experience with RIM/Blackberry 957-era... Great device, completely clueless company regarding an independent software ecosystem.
Say what you will about Apple, but they truly knocked down walls and opened the mobile floodgates to independent developers.
> 0 results
what
I don't want bionic with Surface Flinger. Give me a Wayland compositor and a normal Linux stack.
That's exactly what the Librem 5 is about - https://puri.sm/products/librem-5/
A regular wlroots-based Wayland compositor and regular GNU/Linux apps made adaptive to various screen sizes :)
There's also the PinePhone which is much cheaper because of its lower-end hardware and the fact that its price doesn't include software development.
What drift?
They picked my proposal along with a couple others and gave us some budget to buy hardware and components, and about 3 or 4 weeks to complete the project. It was my first time soldering, first time using an Arduino, and I remember being so stressed out about whether or not it would work.
It all came together in the end and looking back with hindsight now, I really owe a lot to that competition and the phone that let me pretty quickly write python apps on it :) Good times.
The N900 was different. I didn't understand everything I was doing, but installing deb's, overclocking a CPU, installing Android v2.x, installing emulators, browsing forums for more tinkering (I could go on) was the thing that made me love computers more.
Last but not least, it had a killer feature any fresh 18-year-old with friends that had cars loved back then: FM transmitter!! Back when AUX inputs were still a rarity and bluetooth was pairing... pairing... pairing... just pressing a button and tuning to a channel to get crystal clear audio was AMAZING. That and the IR blaster made for some nice party tricks.
The same couldn't be said for N900 or N9, at least for me. J1 was rather slow but good enough, Xperia X was great for a year or so until Android 4.4 got churned, and now with Xperia XA2 and Xperia 10, we are finally there.
However, it's also a painful memory, because I managed to brick the device after spending possibly hundreds of hours customizing it.
The first thing I did was let the micro USB port come loose--a common problem with the device, which seemed to happen to many over time. I remember still being able to charge it somehow (maybe I bought an external charger?). However, the nail in the coffin was when I was cleaning up some "unnecessary" cruft in the init directory, ostensibly to make the thing boot faster.
What I discovered instead was that there was a watchdog that monitored for the existence of crucial files, and, failing to find them, the phone would reboot itself. So my glorious N900 was caught in a bootloop, and moreover without a functioning USB port to re-flash the operating system.
Behind the battery, there are CPU pins that are directly exposed through a grid of copper pads. I made an attempt at wiring up a makeshift soldering job directly to the pins corresponding to a USB port, but failed to confidently pull off the dexterity to work with the small surface, as well as the care to do so without heating things up too much.
Of course I believe there existed a development device with pins that directly clamped onto the pads, but if such a thing was publicly available it was probably hard to come by at affordable prices.
I later tried to realize a similar experience to the N900 on Android using stuff like tmux and F-Droid. However, I have no good memories of any phone after my Nokia, and pretty much don't like phones at all at this point, and now just use my iPhone SE as a glorified flip phone with maps and a web browser.
Genuinely regretting that now.
Running a browser on it was doable, but the bit that really became a blocker for me was the lack of a fast, simple e-mail client (even non-touch screen Blackberries were better, and had slightly better keyboards).
So, several devices originally earmarked for Maemo/Meego (the name changed at some point) actually shipped with Symbian instead.
Some fun facts about Maemo:
Shipped in 2006 with the N770. Had a tablet UI based on X and GTK, a mozilla based browser and was based on Debian. You could build your own packages, add your own apt repositories, and people basically built most of Debian from source to run on it. E.g. I had a JVM running on it and a full LAMP stack at some point.
The N800 was for a time the only device that could run Android (dual boot) and basically shared large parts of the kernel with it. Google used this device for early Android testing and probably owes a lot to Nokia engineers contributing patches to the Linux kernel. The first Nexus phone shipped quite late in the Android development process.
Development of Meego saw several strategy changes that ultimately delayed products for a long time. One of those unfortunate decisions was the decision to swap out GTK for QT; effectively retiring all of the user facing UI. That set back the clock by years. And it got slowed down further by a simultaneous attempt to get Symbian on QT as well. That trainwreck of course ended up sucking up all the resources and failed to ultimately deliver the goods. This was in the middle of Nokia's realization that "oh fuck this iphone thing is real; we need a touch screen UI". It killed at least two touch screen platforms that it had before it came to that realization.
Technically, the N800 was running circles around the first iphone in terms of what you could do on it. All it needed was a bit of hardware polish and a phone stack. Early Android had nothing on Meego. Even the UX was clumsy and it was slow and limited as well. But unlike Meego, Google shipped it and supported it and persisted in developing it. Nokia instead threw out the baby with the bathwater betting on several other horses before ultimately walking away from the whole phone business.
Nokia declined to combine a phone stack + SIM card with the N770 and N800 deeming it to risky; they were protecting their deals with operators to basically cripple phones in favor of crappy operating services. E.g. bundling Skype with the n800 was controversial. It even did video calls with Skype. In 2007. The N800 was a tablet before the iphone, ipad or Android were a thing. There was not a lot else in the market at that point. That thing with a phone stack would have been a killer product despite its many limitations.
The N900 and the later N9 were both labelled as developer phones and cut off from any serious marketing effort. The N9 technically was a decent phone but by the time it shipped, the team had been layed off, the platform cancelled and Nokia was endorsing Windows Phone. It also shipped an Android phone around the same time, leaving the ugly job of cancelling that product to Microsoft after the acquisition.
Samsung's Bada is a direct decendent of Meego. It never really gained any market share against Android and it seems Samsung pretty much gave up on it.
I remember seeing Bada phones at the time, but (at least in my country) the "does it run WhatsApp" era was already starting, and the total lack of apps for Bada made it totally unattractive.
All said, I am not sure that polishing and marketing would have been enough to save the situation. I didn't see Palm's Pre and webOS mentioned in the thread so far: that makes for an example of an excellent platform, with good marketing and capabilities, that still didn't manage to affect the rise of the duopoly.
That would have been a true flagship phone around 2011 with a 12 megapixel camera, aluminium body and nice touch screen. I had it in Symbian form and it was alright but ultimately meh in terms of software. Essentially all Android phones at that point were a combination of slow, not very premium, and riddled with bugs. This was way before Google figured out how to do software updates properly.
A slick UI, linux, touchscreen, etc. exactly what Nokia needed at that point. But they convinced themselves Symbian was good enough and were corrected by the market that made it very clear that it wasn't even close. Of course crippling the device with not enough memory did not help. Nokia saved pennies and sacrificed market share with that.
It's almost hilarious that modern Nokia HMD makes phones now because I expect the company to randomly up and disappear and come back with a WebOS phone or some other weird offshoot.
That was kind of swings and roundabouts, though. The second version of the Maemo app I worked on ended up being a major rewrite Gtk->Qt thanks to the switch, but Qt was so much better.
Seriously, the device was just so attractive both for users and programmers. If Nokia were able to stay the course, they could pull off a bestseller.
Their nosedive into the ground under Stephen Elop is a stuff of dark legends.
The first maemo tablets did not have radio to avoid competion with Symbian, and devices like S90, for example.
Their fate was by no means foreordained and someone different from Stephen Elop (say, Tim Cook?) would have different results.